As AI agents increasingly recommend and purchase products, brands are no longer selling to just one audience. Humans still matter, but a second audience of machines now makes decisions on different criteria, sometimes before a person ever sees a product page. Guest author Toby Coulthard, chief product and growth officer at Jacquard, argues that brands must optimize for both audiences at once rather than relying on the human touch alone.
The first step is to define a distinct brand voice and write it down. In most early-stage companies, voice lives only in the founder's head, which works poorly when the writer or reader is a machine. AI tools follow explicit rules and drift without them. Brands should also document words they never use and claims they never make, since these instincts often feel obvious to teams but go unrecorded.
Product descriptions deserve particular attention. In an agent-mediated world, they may be the most-read brand copy a company owns, because agents rely on them when choosing what to recommend or buy. Two personalized agents may surface the same product for different users with tweaked descriptions, so the source description must carry the brand intact. Brands should also test how AI summarizes their messaging across channels.
Coulthard suggests that startups have an advantage because they can build these practices in from the outset. Established companies must retrofit decades of copy, guidelines and approval processes. Success should be measured through conversions and revenue rather than content volume, preserving human appeal while making the brand legible to machines.
Source: Crunchbase News · Summarized by HeadlinesBriefing